Geometry Based Co-Registration of ALSAT-2A Panchromatic and Multispectral Images

摘要

The co-registration of panchromatic and multispectral images is an important task to obtain a good pan sharpened images. Generally, co-registration can be realized by using image-based or geometry-based methods. The geometry-based method necessitates the understanding and the modelling of the imaging process. This paper describes a method for co-registration of panchromatic and multispectral images from ALSAT-2A the first high resolution Algerian satellite. This method, based on the analysis of the metadata provided by the considered sensor, uses the time dependant collinearity condition for rigorous camera modelling. Three experiments, based on ALSAT-2A images, are performed. The first experiment deals with calculating the exterior orientation parameters (EOP) using the panchromatic image and applying them for georeferencing the multispectral image. In the second experiment; the panchromatic image is georeferenced by applying the model EOP calculated from the multispectral image. The third experiment consists in using both images in a bundle adjustment procedure for a higher accuracy. Sub pixel positioning accuracy can be obtained using the model calculated using the panchromatic image only.